Python

Verified against the Rust reference. Every one of Wickra's 514 indicators is replayed through all 10 languages and checked bit-for-bit against the Rust core's golden fixtures in CI — the math here is provably identical to every other binding (how).

pip install wickra

Zero third-party dependencies — pip install wickra pulls nothing else, not even NumPy. No system dependencies, no C toolchain. The Rust core ships as a pre-built wheel for every Python 3.9–3.13 on Linux, macOS, and Windows (x86_64 and aarch64 where the registry has the platform). NumPy is an optional extra (pip install wickra[numpy]) for zero-copy interop.

• Latest: wickra 0.9.9
• Supported Python: 3.9 / 3.11 / 3.12 / 3.13 (3.10 omitted upstream)

Batch — TA-Lib-style over a series

import wickra as ta                     # zero third-party deps — not even NumPy

prices = [100.0 + i * 0.1 for i in range(1000)]   # a list, array.array or NumPy all work
rsi = ta.RSI(14)
values = rsi.batch(prices)              # array.array('d'), NaN during warmup
                                        # np.asarray(values) wraps it zero-copy if you use NumPy

Streaming — tick-by-tick

import wickra as ta

rsi = ta.RSI(14)
for price in live_feed:
    v = rsi.update(price)               # O(1)
    if v is not None and v > 70:
        print("overbought", v)

Multi-output indicators (MACD, Bollinger …)

import wickra as ta

macd = ta.MACD(12, 26, 9)
for price in live_feed:
    out = macd.update(price)            # dict or None
    if out is not None:
        print(out["macd"], out["signal"], out["histogram"])

bb = ta.BollingerBands(20, 2.0)
out = bb.update(price)                  # {"upper", "middle", "lower", "stddev"}

Candle-input indicators

ATR, Stochastic, and the volume-aware family take a Candle-shaped dict or tuple:

import wickra as ta

atr = ta.ATR(14)
for c in candles:
    v = atr.update({"high": c.h, "low": c.l, "close": c.c})
    if v is not None:
        ...

The dict form is the explicit one; tuples ((o, h, l, c, v, t)) work too and are documented per indicator on the Indicator pages (docs).

Live trading example

The repo ships a runnable example at examples/python/live_binance.py that streams a Binance Spot kline feed through the native wickra.BinanceFeed — no third-party WebSocket client. A wickra.fetch_binance_klines example pulls historical candles over the native REST fetcher the same way.

More

• Quickstart-Python (docs)
• Warmup-Periods — how many bars before each indicator emits a non-None value
• Benchmarks